Key points from article :
A study by Oxford Population Health and Peking University revealed that analyzing blood proteins could improve heart attack risk predictions beyond traditional factors like high blood pressure and genetics. The research analyzed 2,923 proteins in blood samples from 1,976 participants with prior heart attacks and 2,001 participants without, all part of the China Kadoorie Biobank. To ensure broader applicability, the researchers repeated the analysis with data from the UK Biobank, focusing on participants of European descent.
The study identified 446 proteins, such as NT-proBNP and PCSK9, that were strongly associated with heart attacks, even after accounting for conventional risk factors. Adding protein data to existing models significantly enhanced the accuracy of risk predictions. However, individual proteins offered only modest improvements. Using machine learning, researchers narrowed the analysis to 30 proteins that accounted for 90% of the improved prediction in Chinese participants, with comparable results for Europeans.
The findings highlight the potential for blood proteins to serve as predictive tools and treatment targets for heart disease, a leading global cause of death.
"‘This analysis of blood proteins and the role they play in improving risk prediction has highlighted their potential as novel treatment targets for primary heart disease prevention, a leading cause of death and disability around the world," said Dr. Mohsen Mazidi, a genetic epidemiologist at Oxford.
The research was conducted using data from the China Kadoorie Biobank, which enrolled 500,000 adults across China between 2004 and 2008. These findings were presented at the biobank’s 20th-anniversary event in Oxford and published in the European Journal of Epidemiology.
This study showcases advancements in personalized medicine using large-scale population data.
